3DR Solo ESCs Firmware: Updates?
- Thread starter Saijin_Naib
- Start date
Atmega8A
Most of the above is on average correct...
A few points...
It is "Ground Lift"
The Solo ESCs at full power can draw nearly 50A.... that is right... 50A. So 1V lift doesn't seam so bad now does it
The issue is an errata on the ATMega8A...
It sees a TTL high as >2.4V
However it should be >1.4v to be compliant.
So, at 3.3v from the PH2.0, a 1V ground lift is a very bad thing.....
Regarding changing the simonK firmware for Solo...
Simon K himself tuned that firmware to assist as much as possible in this matter...
I wouldn't be tempted to change so something else.
Remember, the exact parameters and calibration are all hard coded just for Solo...
Soo... we're talking about a resistance-based voltage drop, and not a magnetics-induced rise in the local ground potential? Just wanna understand what we're talking about....Atmega8A
Most of the above is on average correct...
A few points...
It is "Ground Lift"
The Solo ESCs at full power can draw nearly 50A.... that is right... 50A. So 1V lift doesn't seam so bad now does it
The issue is an errata on the ATMega8A...
It sees a TTL high as >2.4V
However it should be >1.4v to be compliant.
So, at 3.3v from the PH2.0, a 1V ground lift is a very bad thing.....
Regarding changing the simonK firmware for Solo...
Simon K himself tuned that firmware to assist as much as possible in this matter...
I wouldn't be tempted to change so something else.
Remember, the exact parameters and calibration are all hard coded just for Solo...
A bit of both...
But regardless, neither should be an issue... if the esc was correctly referenced.
But it's not...
So the two possible fixes were...
An opto coupler on the ESC... (best option)
Increasing the voltage of the signal. (Green cube)
- Joined
- Jun 22, 2016
- Messages
- 66
- Reaction score
- 15
- Age
- 51
Sorry, I don't quite understand, but I'm curious. In what situations could the "ground bounce" occur? Not so much in FLY mode but more so in ACRO mode where you are demanding more from the ESC's?
Cheers
Cheers
ehm, TTL logic high is > 2V (for inputs, 2,4V for outputs)
Logic level - Wikipedia
remember an opto coupler would inverse the signal.... i would just use 2 transistors and the necessary pull up resistors
- Joined
- Jun 11, 2017
- Messages
- 30
- Reaction score
- 3
- Age
- 64
Not quite. Before 2.4 Ghz radios, all of the RC receivers seemed to use 5 volt signalling. Some of the Futaba FAAST receivers (and probably others) output 3.3 volt signalling, which did not work correctly with some servos. There exists three wire servo lead connected level convertor hardware to fix this problem.
Chris Shaker
Connecting a module that uses 3.3V signalling to one that uses 5V signalling is a recipe for disaster UNDER NON TYPICAL CONDITIONS. Such as found in any UAV system in the area of motor drive & control. Such as in Space Qualified hardware. Such as in Mil. Spec. hardware.
That's what design reviews are for.
3DR management apparently didn't listen and took the cheap way out.
A hardware fix is the solution to a hardware problem.
We have one with the Green Cube. There may be others, but not yet one that is as Plug & Play.
If this were an avionic system that has to be TSO'd the Green Cube fix would have been mandated by the FAA.
That's what design reviews are for.
3DR management apparently didn't listen and took the cheap way out.
A hardware fix is the solution to a hardware problem.
We have one with the Green Cube. There may be others, but not yet one that is as Plug & Play.
If this were an avionic system that has to be TSO'd the Green Cube fix would have been mandated by the FAA.
A couple of comments:
First, modern RC receiver (or flight controller) output is 3.3 Volts PWM Level. They are usually based on 3.3V CMOS Technology.
There was an issue with some very early Futaba Rx which went down to a level of 2.8V, causing issues with some servo brands (eg Hitec) in combination with long servo cables (voltage drop).
This problem can easily be worked around with a Level Shifter for a few bucks.
Not really. Connecting modules with different PWM signal levels is not an issue at all as long as you do not undercut the defined PWM signal high/low levels as shown in this diagram:
Thats why eg the 5V PWM Level green cube works well with other ESCs and servos.
I was unable to reproduce any issues with Arducopter 3.5 and the stock Solo Cube and stock Solo ESCs over the past 6 months of flying on 2 Solos since the early AC betas. Even under max load conditions in Manual flight modes. If the ground level shift of the Solo ESCs is critical in regard the high/low thresholds (I do not have an oscilloscope at hand to verify this) my ESCs at least seem to be within the tolerances to operate flawlessly ;-)
First, modern RC receiver (or flight controller) output is 3.3 Volts PWM Level. They are usually based on 3.3V CMOS Technology.
There was an issue with some very early Futaba Rx which went down to a level of 2.8V, causing issues with some servo brands (eg Hitec) in combination with long servo cables (voltage drop).
This problem can easily be worked around with a Level Shifter for a few bucks.
Not really. Connecting modules with different PWM signal levels is not an issue at all as long as you do not undercut the defined PWM signal high/low levels as shown in this diagram:
Thats why eg the 5V PWM Level green cube works well with other ESCs and servos.
I was unable to reproduce any issues with Arducopter 3.5 and the stock Solo Cube and stock Solo ESCs over the past 6 months of flying on 2 Solos since the early AC betas. Even under max load conditions in Manual flight modes. If the ground level shift of the Solo ESCs is critical in regard the high/low thresholds (I do not have an oscilloscope at hand to verify this) my ESCs at least seem to be within the tolerances to operate flawlessly ;-)
If you are interested in a procedure that I have found to induce the ESC bug while in flight, send me a PM.
- Joined
- Mar 12, 2016
- Messages
- 3,920
- Reaction score
- 2,623
This loosely translates to "If you want to find out all about crash-o-matic mode, send me a PM"
Would a small level shifter like this at each ESC help, or would the additional possible failure points make it an overall wash?
The pullups would be problematic for the PixHawk 2.0's PWM drivers.
Read this datasheet: http://www.ti.com/lit/ds/symlink/txb0108.pdf
Section 8, page 12 and on (esp. 8.3.5). These are not standard LVC drivers, but "directionless" level translators (which are not doing level translation) and have some special restrictions for their use.
Read this datasheet: http://www.ti.com/lit/ds/symlink/txb0108.pdf
Section 8, page 12 and on (esp. 8.3.5). These are not standard LVC drivers, but "directionless" level translators (which are not doing level translation) and have some special restrictions for their use.
Last edited:
Well, I didn't wanna keep it all to myself. See, some other people wanna crash their precious Solos in the name of science, too.This loosely translates to "If you want to find out all about crash-o-matic mode, send me a PM"
We're all nvts. N-V-T-S, nvts!
no rocket science:
(in an ideal 3DR World - R4 would be in the ESC)
not really sure if there is 5V, but with the appropriate connector this could be plug and play between the Mainboard and the cables to the motors
(in an ideal 3DR World - R4 would be in the ESC)
not really sure if there is 5V, but with the appropriate connector this could be plug and play between the Mainboard and the cables to the motors
Attachments
Last edited:
Or use 4 of these:
http://www.ti.com/lit/ds/symlink/sn74lv1t34.pdf
Cheap too:
SN74LV1T34DCKR Texas Instruments | Integrated Circuits (ICs) | DigiKey
http://www.ti.com/lit/ds/symlink/sn74lv1t34.pdf
Cheap too:
SN74LV1T34DCKR Texas Instruments | Integrated Circuits (ICs) | DigiKey
@just_bruce please let me know how you achieved this. I wasn't able to break the ESCs so far with 3.5 and my parameter settings. Maybe I have overseen something ;-)
Similar threads
New Posts
-
-
-
-
Free Music / SFX Resource for Your Videos - Over 2000 Tracks- Latest: Eric Matyas
-
Members online
No members online now.